Yellow Flags

The usefulness of specimen data—particularly its usefulness to non-botanists or to analytical software—may be compromised by data errors. Therefore, we are flagging records that can help us resolve possible errors in the Jepson eFlora or data associated with CCH records.

Specimen records that are shown with a yellow flag may not have any errors associated with the specimen. In these cases, the yellow flags indicate that a change is required in the distribution string of the Jepson eFlora. Range discrepancies have been caused by oversights, by failure to incorporate the consequences of synonymy, and (especially for non-native taxa) by ongoing range expansion. These are false-positives which need to be corrected in the Jepson eFlora.

For example, consider the taxon Salvia mohavensis. According to the Jepson eFlora treatment for S. mohavensis, it occurs in DMtns (Desert Mountains) and DSon (Sonoran Desert). However, when a map of CCH specimens of S. mohavensis is generated, the majority of specimens are distributed throughout the eastern Mojave Desert. In this case the CCH data are not erroneous; the Jepson eFlora distribution string must be updated to include DMoj.


Left: Map of bioregions in which Salvia mohavensis occurs, according to the Jepson eFlora. Right: Map of CCH specimens of S. mohavensis with points mapping outside the asserted bioregion boundaries colored yellow.

Yellow flags may also be generated for other reasons:

Incorrect coordinates
Latitude or longitude might be incorrectly recorded, converted, or transcribed. The location of the record might have been recorded incorrectly or inferred erroneously. These can often be corrected by examining the data without looking at a specimen.
Mapping errors or other software errors
Very small region units such as pieces of DMtns or offsh ore islets were not represented well by the process and consequently result in false positives. Records with locations close to boundaries are troublesome.
Misidentification or other naming issues
The specimen might have been incorrectly keyed out or mismatched with a picture initially, or there might have been a taxonomic revision subsequent to the initial identification. Assessing the possibility of misidentification will usually involve examination of specimens. It will be dangerous in most cases to make inferences based on record data alone.

If you want to help investigate and resolve these issues, please contact Staci Markos (smarkos@berkeley.edu) or David Baxter (dbaxter@berkeley.edu).

Names with 100+ yellow flags


Here's an interesting one